There was the "what license is that post actually under? Is it 2.5? 3.0? 4.0?" which made things difficult.
There was the "what are the actual attribution requirements that SE has for sites that use its content?" This is a bit of an issue because it's never really clear what those requirements are and what you need to do. It can also hurt SEO because it's duplicated content. Furthermore, codidact leadership had already and enough dealings with SE lawyers and likely wanted to avoid any other.
Lastly, there was the desire to make a philosophical break with SE. The codidact founders didn't want to have anything to do with SE.
Some sites are doing ok. Others stood up but didn't have sufficient involvement to keep them going.
For a counter example, "PhysisOverflow" has an import tool that they use. https://www.physicsoverflow.org/4536/import-queue
Having an imported site that is mostly inactive with activity on that same content is even more disappointing than having a mostly empty site. And active mirroring is a time-consuming process that runs into rate limit issues with an API.
https://software.codidact.com/categories/38
I haven't used codidact (sorry, name needs replacing), ok just poked around
* too slow
* needs type ahead find search
* needs a GIST experience
The site looks good, presentation is really clean. Lots to like about it. But the think that replaces SO is going to have to be a step function in capabilities. That said, just fixing the weird descend into performative rule following and language-lawyering on SO might be that step function. * "Tipping" or actually giving money to a question answerer would be cool
* Having a question asker being able to put a bounty on question would be cool
On the face of it, I am not getting scalability (in many senses) vibes from codidact.